Information Technology Reference
In-Depth Information
Fig. 18.4
Linear and cubic fittings to the regression model
Syst
vs.
Bmi
Table 18.4
Cluster centers (fuzzy rules) for the regression model
Syst
vs.
Bmi
Rule If
Bmi
then
Syst
1
20.5
120
2
22.5
145
3
27.4
160
4
27.5
135
5
29.1
190
6
33.9
2
fitting plane or surface which is optimally close to our data points. Table 18.5 shows
the linear resolution and thus we obtain the regression equation
Syst = 0.856
Age
+ 3.524
Bmi
+ 15.421.
According to t-tests in Table 18.5 , both independent variables seem to be relevant
in our model. The RMSE is 18.87 (adjusted Rsquare = .432) which smaller than in
the foregoing linear model Syst vs. Bmi and almost similar to our fuzzy model with
one independent variable. Figure 18.8 depicts this linear fitting.
